1. The same wires are at both ends of the cable.
2. The same connector components are at both ends of the cable.
3. It's like having rigging assembly that picks up a a large load. A chain is no stronger at one end than it is at the other. When we install underground concrete structures, they are lifted with a rigging chain which connected to a ring that has 3 chains. All chains are the same size and same rating ... if the rigging apparatus chains are rated say at 10,000 pounds each, the three chains below the ring do not increase the capacity of the single chain from the ring to the machine doing the lifting. You can likely lift 5 times the "rated load" before that chain breaks ... but don't do it with a safety inspector on site. Those chains are subject to dings, notches, corrosion , repetitive load cycling all of which decreases strength over time and relying on break strength will eventually get someone killed.
So the issue is rating versus ultimate capacity. And the rating is below capacity for a reason:
1. All but the cheapest PSUs can exceed their rating .... but it's not recommended
2. Intel CPUs can go up to just under 100C .... but it's not recommended
3. nVidia GPUs have max temps in the neighborhood of 95C .... but it's not recommended... nVidia starts throttling at 82C
4. Every cable, outlet, switch and electrical device known to man has a rating below it's ultimate capacity. When designing power service to a building, if I was to ignore ratings and design based upon failure points, I'd be in violation of all applicable codes and I would lose my license to practice.
So yes, ratings are not intended to convey the maxium capacity but, as you said ... "for universal safety issues due to heat output because full, solid connections can't be guaranteed with spreadable pins subject to stresses and wear"
I have a Seasonic cable like that ... except it has more than 8 cables/connector points at the PSU end.